[GRASS-SVN] r49391 - in grass-addons/raster/r.pi: r.pi.corrwin
r.pi.csr.mw r.pi.energy r.pi.nlm r.pi.searchtime
svn_grass at osgeo.org
svn_grass at osgeo.org
Sun Nov 27 10:23:54 EST 2011
Author: wegmann
Date: 2011-11-27 07:23:54 -0800 (Sun, 27 Nov 2011)
New Revision: 49391
Modified:
grass-addons/raster/r.pi/r.pi.corrwin/description.html
grass-addons/raster/r.pi/r.pi.csr.mw/description.html
grass-addons/raster/r.pi/r.pi.energy/description.html
grass-addons/raster/r.pi/r.pi.nlm/description.html
grass-addons/raster/r.pi/r.pi.searchtime/description.html
Log:
extened manual pages, added various working examples
Modified: grass-addons/raster/r.pi/r.pi.corrwin/description.html
===================================================================
--- grass-addons/raster/r.pi/r.pi.corrwin/description.html 2011-11-27 14:55:44 UTC (rev 49390)
+++ grass-addons/raster/r.pi/r.pi.corrwin/description.html 2011-11-27 15:23:54 UTC (rev 49391)
@@ -1,35 +1,23 @@
<h2>DESCRIPTION</h2>
This module computes the correlation between two raster files but unlike
-<em>r.pi.lm</em> for moving windows of a specific size. This module used
-code sniplets from <em>r.neighbours</em> and <em>r.covar</em>.
+<em>r.pi.lm</em> within a moving window of a specific size. This module is partly based on <em>r.neighbours</em> and <em>r.covar</em>.
<h2>EXAMPLE</h2>
An example for the North Carolina sample dataset:
+Correlation of all pixels within a 7x7 sized window of two rasters (elevation and slope). The output is multiplied by 10000 for
<div class="code"><pre>
-g.region -d
-...
+r.pi.corrwin input1=slope input2=elevation output=corrwin1 size=7 max=10000
+r.colors corrwin1 col=bgyr
</pre></div>
<h2>SEE ALSO</h2>
<em>
-<a href="r.pi.corearea.html">r.pi.corearea</a>,
-<a href="r.pi.csr.mw.html">r.pi.csr.mw</a>,
-<a href="r.pi.export.html">r.pi.export</a>,
-<a href="r.pi.graph.html">r.pi.graph</a>,
-<a href="r.pi.graph.dec.html">r.pi.graph.dec</a>,
-<a href="r.pi.graph.iter.html">r.pi.graph.iter</a>,
-<a href="r.pi.graph.red.html">r.pi.graph.red</a>,
-<a href="r.pi.grow.html">r.pi.grow</a>,
-<a href="r.pi.import.html">r.pi.import</a>,
<a href="r.pi.index.html">r.pi.index</a>,
<a href="r.pi.lm.html">r.pi.lm</a>,
-<a href="r.pi.odc.html">r.pi.odc</a>,
-<a href="r.pi.prob.mw.html">r.pi.prob.mw</a>,
-<a href="r.pi.rectangle.html">r.pi.rectangle</a>,
<a href="r.pi.html">r.pi</a>
</em>
Modified: grass-addons/raster/r.pi/r.pi.csr.mw/description.html
===================================================================
--- grass-addons/raster/r.pi/r.pi.csr.mw/description.html 2011-11-27 14:55:44 UTC (rev 49390)
+++ grass-addons/raster/r.pi/r.pi.csr.mw/description.html 2011-11-27 15:23:54 UTC (rev 49391)
@@ -1,6 +1,6 @@
<h2>DESCRIPTION</h2>
-Complete Spatial Randomness analysis on moving window.
+Complete Spatial Randomness (CSR) analysis based on a moving window. This function uses either the Clark and Evans (1954) or Donnelly (1978) aggregation index for testing of clustering of point patterns.
<h2>NOTES</h2>
@@ -10,11 +10,20 @@
An example for the North Carolina sample dataset:
+Compute the CSR for the whole landscape of <em>landclass96</em> using class 5 (1000 iteration) using the Clark Evans method:<br>
<div class="code"><pre>
-g.region -d
-...
+v.random output=randompoints n=100 zmin=0.0 zmax=0.0
+v.to.rast input=randompoints output=randompoints use=val val=1
+r.pi.csr.mw input=randompoints keyval=1 n=1000 method=clark_evans output=csr1
</pre></div>
+The results for the whole landscape is prompted to the console.
+Compute the CSR for a defined moving window size of <em>landclass96</em> using class 5 (1000 iteration, Clark Evans method):<br>
+<div class="code"><pre>
+r.pi.csr.mw input=randompoints keyval=5 n=1000 method=clark_evans size=7 output=csr1
+</pre></div>
+
+
<h2>SEE ALSO</h2>
<em>
Modified: grass-addons/raster/r.pi/r.pi.energy/description.html
===================================================================
--- grass-addons/raster/r.pi/r.pi.energy/description.html 2011-11-27 14:55:44 UTC (rev 49390)
+++ grass-addons/raster/r.pi/r.pi.energy/description.html 2011-11-27 15:23:54 UTC (rev 49391)
@@ -1,6 +1,6 @@
<h2>DESCRIPTION</h2>
-Individual-based dispersal model for connectivity analysis - energy based.
+Isolation or connectivity of singular patches of a defined landcover class using individual-based dispersal models. This function is based on a maximum amount of energy for each individuals dispersing through the landscape which is deminished by a fricition or cost map.
<h2>NOTES</h2>
@@ -11,6 +11,8 @@
An example for the North Carolina sample dataset:
<div class="code"><pre>
+
+
g.region -d
...
</pre></div>
Modified: grass-addons/raster/r.pi/r.pi.nlm/description.html
===================================================================
--- grass-addons/raster/r.pi/r.pi.nlm/description.html 2011-11-27 14:55:44 UTC (rev 49390)
+++ grass-addons/raster/r.pi/r.pi.nlm/description.html 2011-11-27 15:23:54 UTC (rev 49391)
@@ -4,19 +4,26 @@
<h2>NOTES</h2>
-Related to r.pi.nlm but using fractal landscapes instead of circular growth. Per default the size of the whole region is used for generating a random landscape, this can be constraint by assigning a class in the raster map with acts as mask for the generation of the random landscape (<em>nullval</em>).
+Related to r.pi.nlm.circ but using fractal landscapes instead of circular growth. Per default the size of the whole region is used for generating a random landscape, this can be constraint by assigning a class in a raster map with acts as mask for the generation of the random landscape (<em>nullval</em>).
The landcover can be set manually, randomly or be taken from the input class coverage. The agglomeration level (<em>sharpness</em>) can be set manually or randomly. If similar random landscape with differing e.g. percentage coverage should be generated, then the <em>seed</em> can be set using any number and reused for any subsequent analysis.
<h2>EXAMPLE</h2>
An example for the North Carolina sample dataset:
-A random landscape is generated using the percentage coverage of class 5. The agglomeration factor is set randomly.<br>
+A random landscape with random percentage coverage and agglomeration factor:<br>
<div class="code"><pre>
-<b>r.pi.nlm input=</b>landclass96 <b>output=</b>nlm.1 <b>keyval=</b>5 <b>--o</b>
+r.pi.nlm output=nlm.1 landcover=50 --o
</pre></div>
+
+A random landscape is generated using the percentage coverage of class 5. The agglomeration factor is set randomly:<br>
+<div class="code"><pre>
+r.pi.nlm input=landclass96 output=nlm.2 keyval=5 --o
+
+</pre></div>
+
<h2>SEE ALSO</h2>
<em>
Modified: grass-addons/raster/r.pi/r.pi.searchtime/description.html
===================================================================
--- grass-addons/raster/r.pi/r.pi.searchtime/description.html 2011-11-27 14:55:44 UTC (rev 49390)
+++ grass-addons/raster/r.pi/r.pi.searchtime/description.html 2011-11-27 15:23:54 UTC (rev 49391)
@@ -1,26 +1,76 @@
Individual-based dispersal model for connectivity analysis (time-based)
+
<h2>DESCRIPTION</h2>
-This modules aims at generating sampling areas which are only known by the
-coordinate of one corner. The input are single points, while the output are
-areas representing the corresponding area for each of the single
-points/coordinates.
+This modules provides information about the isolation or connectivity of individual fragments derived of a landcover classification. Unlike <em>r.pi.energy</em> this module provides information about the time from emigration to immigration. The individual based dispersal model results are based on the step length and range, the perception distance and the attractivity to move towards patches.
<h2>NOTES</h2>
-The areas can only be generated horizontally, not diagonal. This can be added
-as wish and might be implemented in the future.
<h2>EXAMPLE</h2>
An example for the North Carolina sample dataset:
+
+The connectivity of patches of the <em>landclass96</em> class 5 are computed using the time from emigration to immigration. The step length is set to 5 pixel, the output statistics are set to <em>average</em> time and <em>variance</em> of searchtime. For each patch 1000 individuals were released and the model stopped when at least 80% of all individuals sucessfully immigrated:<br>
<div class="code"><pre>
-g.region -d
-...
+r.pi.searchtime input=landclass96 output=searchtime1 keyval=5 step_length=5 stats=average,variance percent=80 n=1000
</pre></div>
+constrain the angle of forward movement to 10 degrees:
+<div class="code"><pre>
+r.pi.searchtime input=landclass96 output=searchtime1 keyval=5 step_length=5 stats=average,variance percent=80 n=1000 step_range=10
+</pre></div>
+
+setting the perception range to 10 pixel:
+<div class="code"><pre>
+r.pi.searchtime input=landclass96 output=searchtime1 keyval=5 step_length=5 stats=average,variance percent=80 n=1000 perception=10
+</pre></div>
+
+setting the perception range to 10 pixel:
+<div class="code"><pre>
+r.pi.searchtime input=landclass96 output=searchtime1 keyval=5 step_length=5 stats=average,variance percent=80 n=1000 perception=10
+</pre></div>
+
+increasing the attraction to move towards patches to 10:
+<div class="code"><pre>
+r.pi.searchtime input=landclass96 output=searchtime1 keyval=5 step_length=5 stats=average,variance percent=80 n=1000 multiplicator=10
+</pre></div>
+
+increasing the attraction to move towards patches to 10:
+<div class="code"><pre>
+r.pi.searchtime input=landclass96 output=searchtime1 keyval=5 step_length=5 stats=average,variance percent=80 n=1000 maxsteps=10
+</pre></div>
+
+output of each movement location for a defined step frequency. Here every 10th step is provided as output raster:
+<div class="code"><pre>
+r.pi.searchtime input=landclass96 output=searchtime1 keyval=5 step_length=5 stats=average,variance percent=80 n=1000 out_freq=10
+</pre></div>
+
+output of a raster which immigration counts:
+<div class="code"><pre>
+r.pi.searchtime input=landclass96 output=searchtime1 keyval=5 step_length=5 stats=average,variance percent=80 n=1000 out_immi=immi_counts
+</pre></div>
+
+output of a binary immigration matrix. Each patch emigration and immigration for all patch combinations is recorded as 0 or 1:
+<div class="code"><pre>
+r.pi.searchtime input=landclass96 output=searchtime1 keyval=5 step_length=5 stats=average,variance percent=80 n=1000 binary_matrix=binary_matrix.txt
+</pre></div>
+
+output of a matrix with immigration counts for each patch:
+<div class="code"><pre>
+r.pi.searchtime input=landclass96 output=searchtime1 keyval=5 step_length=5 stats=average,variance percent=80 n=1000 immi_matrix=immi_counts.txt
+</pre></div>
+
+the previous examples assumed a homogeneous matrix, a heterogenous matrix can be included using a raster file which values are taken as costs for movement (0-100):
+<div class="code"><pre>
+# it is assumed that our species is a forest species and cannot move through water, hence a cost of 100, does not like urban areas (class: 6, cost: 10) but can disperse through shrubland (class 4, cost=1) better than through grassland (class 3, cost: 2):
+r.mapcalc "cost_raster = if(landclass96==5,0,if(landclass96 == 1, 10, if (landclass96==3,2, if(landclass96==4,1,if(landclass96==6,100)))))"
+r.pi.searchtime input=landclass96 output=searchtime1 keyval=5 step_length=5 stats=average,variance percent=80 n=1000 costmap=cost_raster
+</pre></div>
+
+
<h2>SEE ALSO</h2>
<em>
More information about the grass-commit
mailing list